Wondering if there is better interpolation method to use for water quality sampling points.

[JessicaRutherford2]

I having been creating interpolation surface maps, from set water quality sample points, and have been using the Spatial analysis tool>interpolation>IWD to get my interpolated surfaces. Macrophyte (water quality samples) data is recorded in ug/m2. I  have a set of categories I use in symbology tab to show the various measured categories (as seen in my images below). However, some sample points only show low levels of macrophytes at specific sampling sites, but when I use the IWD interpolation and the set categories it has the whole area showing that it is a low range. I am not too sure if this can be corrected by changing my categories or if there is a better interpolation method to use. 

Any suggestion or advice would be great. 

[JayantaPoddar]

There are different Interpolation methods that you could try. Could you also label the points with the Field that you are using to interpolate, so that we can get an idea about the distribution of values?

[DanPatterson_Retired]

The interpolation method is one thing, but as is shown, the symbology is another. Your pattern of points is a challenge at best, but you have initially constricted the ability to display what is there by your imposition of the classification scheme, which I assume has some basis for its choice.

If you have the Geostatistical Analyst, then there are other interpolation methods.  slynch-esristaff‌ might have some good suggestions given your pattern of points and potentially the values obtained at those points.

[SteveLynch]

As Dan Patterson‌ mentions -> there are other interpolation methods. I would suggest that you look at Spline with Barriers (Spatial Analyst/3D) and Kernel Interpolation with Barriers (Geostatistical Analyst).

Also, as jayanta.poddar‌ mentions, display the value at the points and take a screenshot only showing the interpolated area.

[JessicaRutherford2]

Thanks Steve

I will look at these two methods and see what works. I am still a little skeptical that I can do anything different, as I am limited to the sampling points and the classification I am required to use. 

Also, I am not too sure what you mean by...'display values at the points'..? More like I am not too sure how to do this. 

Thanks for your input.

Very appreciated. 

Jessica 

[JessicaRutherford2]

Thank Dan

Yes, I realize I am limited with how to display this, but my client after four years is asking if I it can be a bit more realistic. I suggest more sampling sites, but that is not possible, and they want to stay with the sample classification scheme. Thanks for your input. I will research Geostastical Analyst for other interpolations methods. cheers jess

[DanPatterson_Retired]

You should explore the interpolator first... then worry about the classification scheme, or at least show an alternate if the first one is unrealistic in your estimation.  In fact, post your results back for commentary.  

Steve means to label the sampling points with the 'z' value (what you used to interpolate) to give some idea what the actual values are at the points so that they can be examined in light of the interpolation results.

[JessicaRutherford2]

Hi Dan

Thanks. I see what you mean now. I've worked out the labels for sampling points. I will try the suggested interpolation and post back for comments and suggestion tomorro. Thanks for the help.

Cheers

Jess